[LLVMdev] FW: RFC: Supporting different sized address space arithmetic
Micah.Villmow at amd.com
Thu Sep 6 09:19:00 PDT 2012
Doh! hit send too soon, patch attached.
> -----Original Message-----
> From: Villmow, Micah
> Sent: Thursday, September 06, 2012 9:17 AM
> To: 'Eli Friedman'
> Cc: LLVM Developers Mail
> Subject: RE: [LLVMdev] FW: RFC: Supporting different sized address space
> Here is the first of many patches that adds support for specifying
> different pointer sizes for different address spaces.
> This is only the modifications to TargetData and not all the changes to
> the backends/optimizers. There should be no functional changes here
> since the default value is what the current value is.
> After this is approved, my goal is the following:
> 1) Add a few interfaces to various functions that simplify retrieving
> address space information.
> 2) Update all of the optimizations to use address space information when
> querying the pointer type.
> 3) Update the backends to follow suite
> 4) Update the clients(clang, etc..) to use the correct API.
> 5) remove the default arguments so that future users must explicitly
> specify an address space.
> I'm not sure how to add tests for this since no backend uses it yet, but
> i'll try to figure something out.
> > -----Original Message-----
> > From: Eli Friedman [mailto:eli.friedman at gmail.com]
> > Sent: Thursday, August 30, 2012 3:43 PM
> > To: Villmow, Micah
> > Cc: LLVM Developers Mail
> > Subject: Re: [LLVMdev] FW: RFC: Supporting different sized address
> > space arithmetic
> > On Thu, Aug 30, 2012 at 3:27 PM, Villmow, Micah
> > <Micah.Villmow at amd.com>
> > wrote:
> > >
> > >
> > >> -----Original Message-----
> > >> From: Eli Friedman [mailto:eli.friedman at gmail.com]
> > >> Sent: Thursday, August 30, 2012 3:03 PM
> > >> To: Villmow, Micah
> > >> Cc: LLVM Developers Mail
> > >> Subject: Re: [LLVMdev] FW: RFC: Supporting different sized address
> > >> space arithmetic
> > >>
> > >> On Thu, Aug 30, 2012 at 2:38 PM, Villmow, Micah
> > >> <Micah.Villmow at amd.com>
> > >> wrote:
> > >> > Eli,
> > >> > Here is an updated patch. This is a lot smaller based on your
> > >> feedback and still solves the same problem.
> > >>
> > >> The patch appears to be corrupt; can you regenerate it?
> > > [Villmow, Micah] Attached.
> > >>
> > >> > For your comment on the IR changes, I'm reluctant to introduce
> > >> > changes
> > >> there because really the backend is overriding the default behavior
> > >> at a device specific level. The optimizations themselves can be
> > >> dangerous, but still should produce correct results, this only
> > >> allows the backend to optimize certain cases and is opt-in.
> > >>
> > >> Suppose I have an array of ten pointers into some address-space
> > >> which uses 16-bit pointers, and the default pointer size is 64
> > >> bits. How many bytes in memory does that take? To me, it seems
> > >> like the obvious answer is 20 bytes, but if you compute it using
> > >> our current TargetData, you'll come up with an answer of 80. That
> can't work.
> > >>
> > >> If your answer is that it should be 80, and the size of a pointer
> > >> isn't something the frontend/IR optimizers should be aware of, I'm
> > >> not sure your approach makes sense; you could just introduce custom
> > >> load/store nodes in your target which truncate the pointer, and you
> > >> wouldn't need to mess with the size of a pointer at all.
> > > [Villmow, Micah] Yeah I see your point here. I don't deal with array
> > of pointers in OpenCL, so didn't think of this case.
> > >
> > > So, what about extending data layout to support something like:
> > > p#:size:abi:pref <- specify the size, abi and preference for the
> > pointer of address space '#'. Default behavior is p:size:abi:pref.
> > That's fine.
> > (You'll also need to deal with the fact that LLVM assumes bit casts
> > across address-spaces are lossless.)
> > -Eli
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 17181 bytes
More information about the llvm-dev